When performance shortfall arises, contract or trust? A multi-method study of the impact of contractual and relational governances on performance in public – private partnerships

(1)

Tilburg University

When performance shortfall arises, contract or trust? A multi-method study of the impact of contractual and relational governances on performance in public – private partnerships Fang, F. DOI: 10.26116/center-lis-1932 Publication date: 2019 Document Version

Publisher's PDF, also known as Version of record Link to publication in Tilburg University Research Portal

Citation for published version (APA):

Fang, F. (2019). When performance shortfall arises, contract or trust? A multi-method study of the impact of contractual and relational governances on performance in public – private partnerships. CentER, Center for Economic Research. https://doi.org/10.26116/center-lis-1932

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal

Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

(2)

When performance shortfall arises, contract or trust?

A multi-method study of the impact of contractual and relational governances on performance in Public – Private Partnerships

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan Tilburg University

op gezag van de rector magnificus, prof. dr. K. Sijtsma,

in het openbaar te verdedigen ten overstaan van een door het college voor promoties aangewezen

commissie

in de aula van de Universiteit

op vrijdag 8 november 2019 om 10.00 uur door

Feng Fang

(3)

ii Promotores:

prof. dr. ir. H.A. Akkermans prof. dr. ir. G.C.J.M. Vos Copromotor:

dr.ir. W. van der Valk

Overige leden van de Promotiecommissie: prof. dr. J. Roehrich

prof. dr. J.Y.F. Wynstra

(4)

iii

Acknowledgement

This PhD research has been a long, exciting and life-changing journey for me and it would not have been made possible without the help, support and advice that I received from many people.

I would like to give my first thanks to my supervisors, Henk Akkermans, Bart Vos and Wendy van der Valk. Without their guidance, feedback and encouragement, this PhD would not have been achievable. I feel so happy and lucky that I have had such a great supervisor team. Henk, you are the one who always has so many innovative ideas. Bart, you are the one who is able to reshape these new ideas and make them practically feasible for me. And Wendy, you are the one who is being critical and strict, but in a good way to challenge and help me achieve better performance. Special thanks to Henk and Bart for not only being my advisors in research, but also in life. There were many difficult moments during my PhD, and it was their special way of encouragement (with jokes) that helped me out. I would like to thank Henk also for inspiring me to become an entrepreneur.

I would also like to give my gratitude to the members of the PhD committees, Jens Roehrich, Finn Wynstra, Leo van Dongen, Niels Noorderhaven, and Erik de Ridder for the challenging questions and also the constructive suggestions during my predefense. I have greatly benefited from the insights provided by them.

(5)

iv

Holtzer, Ingeborg Leermakers-Doppenberg, André Smits, Joost Bekkers, Frank Verkuijlen, Har Coenen, Johan Gerritsen, Roger Janssen, Ad de Man, Wil Pörteners, Giel Geraeds, Wim Sanders, Stefan van der Linden, Frans Franken, Frans van der Molen, Ivo Dellemann, Antoine van Geffen, Bart van de Ven, Kasper Nipius, Kees Monsieurs, Henk Schoones, Frank van Zijl, Roland Boer, Rini van Haperen, Edwin den Heijer and Peter van Gurp. Also many thanks go to people from private contractors, who sacrificed their precious time and have been very cooperative for my interviews. For confidentiality, here I only mention their first names, Maurice, Johan, Marc, Kevin and Jan.

Special thanks go to my research assistant Kees van der Linden. Kees, I enjoyed a lot working with you and your work has contributed greatly to this research.

To all my deal friends and colleagues at Tilburg University, thank you very much for sharing so many unforgettable moments, whisky nights, dinners in Tilburg city center, BBQ nights, beach volleyball tournaments, coffee breaks at P1.137, lunch breaks at Mensa, beers at Esplanade, walks in the forest, weekends in Germany & Belgium …. Too many to name them all. Thank Roland and Laurens for being my business partner & colleague, friends and paranymphs. I would like to thank Roland also for motivating me to finish my PhD as soon as possible by having a bet with me, although he knew he was very likely to lose.

A very special thank goes to my Dutch teacher and friend, Marcel Noordermeer. After a busy working day, a 90-minute conversation in Dutch can really calm me down. I would also like to say a heartfelt thank you to my mum, dad and mother-in-law for encouraging me to pursue a PhD degree, although none of them have a clue on what kind of research I actually do.

Finally to my loves, Juan and Yilin.

(6)

v

Introduction

(15)

2

are discussed since each plays an important role in determining the performance dynamics of the projects. Many discussions (e.g. van der Valk, Sumo, Dul, & Schroeder, 2016) have focused on what the correct governance might be for innovation projects. However, there has been limited research investigating how the choice of governance mechanisms influences the performance dynamics, in particular when there is a performance shortfall. This research aims to fill this gap by addressing the following research question: “how do two governance strategies for dealing with performance shortfall and performance affect each other over time in an outsourced infrastructure innovation project that is a PPP?” In Section 1.4, the structure of the dissertation is outlined.

1.1 Research context

The infrastructure sector

1.1.1.1 Importance of the infrastructure sector

The infrastructure sector consisting of such vital elements as water supply, electricity supply, gas supply, sanitation, telecommunication, railway and road transport, is of critical importance for our civilized society. This infrastructure is the backbone of reliable, affordable and qualified services that ensure a livable, accessible and safe society. As Finger et al. (2005) point out, an infrastructure consists of complex technical, economic, and political systems provides essential services to the whole of society. 1.1.1.2 Characteristics of infrastructure systems

(16)

3

for almost 80 years without major re-investment. Power lines last for about 40-60 years and sewers can reach a useful life of 80 years before they have to be replaced. The third feature is the dominance of public organizations in the infrastructure sector. These organizations may be corporations with public shareholders, but also organizations that are directly controlled by public organizations. The fourth feature relates to the high degree of regulation, that for example includes price regulation, service quality norms, rules for network access and environmental regulations. Finally, the fifth feature of infrastructure is the high degree of “systemness”, implying that there are strong complementarities among system components, which in turn means that overall system performance is highly dependent on how well the various components are coordinated (Künneke, 1999). This strong interdependency is especially present in infrastructure like electricity supply, railway transport and telecommunication. In other network sectors, such as road transportation, these interdependencies are less prominent. 1.1.1.3 The Dutch infrastructure sector

(17)

4

and waste-water purification. They manage more than 3,400 pumping stations, 230,000 km of drainage ditches and around 350 sewage treatment plants. They also manage around 3,450 km of primary flood defense structures and 14,000 km of secondary flood defense structures (Havekes et al., 2017). All these examples of infrastructure are important in safeguarding a livable, accessible and safe Netherlands.

1.1.1.4 Demand for cost efficiency and innovation

In the past, many types of infrastructure were not maintained or renewed adequately, which led to increasing failures and an accumulation of investment needs (Gil & Beckman, 2009). While public organizations are under pressure to achieve cost efficiency, the infrastructure sector is highly regulated. Public organizations have to manage assets governed by many rules and regulations, with increasing attention being paid to environmental issues. For example, in the case of electricity or gas supply, depletion of fossil fuel resources, air pollution or nuclear waste are examples of negative effects associated with energy consumption. In the case of waste-water treatment, reduction of phosphate and nitrate burdens on surface waters is one of the key objectives of water authorities. Innovation is needed to deal with these environmental issues. In conclusion, the challenge for many public organizations is to build and operate infrastructure more cost efficiently while at the same time becoming more innovative.

Public – private partnerships (PPPs)

(18)

5

the project), and a payment structure which is influenced to some degree by service volumes and/or service quality.

1.1.2.1 Benefits of PPPs

According to Verweij (2015), a PPP has many benefits in infrastructure development: long-term benefits, such as “value-for-money” (Grimsey & Lewis, 2004; Kwak, Chih, & Ibbs, 2009), and short-term benefits, including a reduction in cost and time taken to deliver infrastructure services, higher quality service delivery, lower administrative costs, and the transfer of risks to the private sector (Flyvbjerg, Skamris, & Buhl, 2003; Hodge & Greve, 2007; Kwak et al., 2009). To elaborate more, there are three main advantages of PPPs (Klijn, Koppenjan, & De Boer, 2014). First, a PPP helps to secure the same outcomes for lower costs through cooperation (efficiency). Second, it brings greater outcomes for the same cost (added value). Public and private actors can add value to each other’s performance because their efforts enhance the value of the product or service that is being delivered. Third, it results in innovative solutions, or solutions that have not been achieved before, by harnessing each other’s knowledge and expertise (Borys and Jemison, 1989; Hodge and Greve, 2005).

1.1.2.2 PPPs in the Netherlands

(19)

6

contracts, with non-performance or underperformance resulting in no payment or the application of penalties. This performance-related payment is also one of most important characteristics of Performance Based Contracts (PBCs) (Selviaridis & Wynstra, 2015), which focus on specification and evaluation of outputs/outcomes rather than required inputs, activities or processes (Martin, 2007). During the contracting process, performance is determined by a monitoring system designed by the private party. In 2016, 37 national infrastructure projects were realized using DBOM contracts in the Netherlands, with a total value of 13 billion euros (Koppenjan & de Jong, 2017). The government has implemented PPP contracts on the presupposition that it would, on one hand, be able access more innovative solutions and, on the other, to complete large infrastructure projects faster and more efficiently (Van Valkenburg & Nagelkerke, 2017).

1.2 Research problem: performance shortfalls

(20)

7

(21)

8

1.3 Research question

(22)

9

aims to investigate how different governance mechanisms (i.e. contractual and relational governance mechanisms) and performance affect each other over time in a public – private partnership (PPP).

As the study unfolded, I discovered that in all cases the two parties in a PPP encountered a performance shortfall. When it occurs, the business relationship is slightly changed, it deteriorates and its outcome is uncertain (Vidal, Fenneteau, & Paché, 2016). It can give rise to dissolution or cause reactions that lead to the restoration of relationships. The two parties’ different strategies for dealing with a performance shortfall are highly relevant in determining project performance and the relationship between partners over time, and in turn project performance and relationship also affect how different strategies are used. Thus, I extended the literature review into the domain of conflict management, subject to the theoretical framework of governance mechanisms. I discovered that many existing endeavors have been devoted to studying conflict development and escalation (Bijlsma-Frankema, Sitkin, & Weibel, 2015; Coleman, Vallacher, Nowak, & Bui-Wrzosinska, 2007; Perlow & Repenning, 2009), however, seldom has the theoretical lens of governance mechanisms been applied to the domain of conflict research (Lumineau and Henderson, 2012 and Malhotra and Lumineau, 2011 are two exceptions). To fill this gap, I further refined the research question as “how do two governance strategies for dealing with performance shortfall and performance affect each other over time in an outsourced infrastructure innovation project that is a PPP?”

1.4 Structure of the dissertation

(23)

10

conflict resolving strategies, which relate to the research problem. After that the research question is refined, the subsequent four chapters (Chapter 4-7), present narratives of the four cases, namely those of Riverboard – Mouse, Brookboard – Elephant, Creekboard – Giraffe and Lakeboard – Panda, and reflect on narratives from static and dynamic perspectives. Chapter 8 shows the findings from cross-case analysis under a static perspective, while Chapter 9 demonstrates findings under a dynamic perspective. Subsequently, Chapter 10 uses a system dynamic simulation model to explain and support the theoretical propositions. Finally, Chapter 11 concludes the dissertation with the main findings, theoretical contributions, managerial applications, limitations and directions of future research.

Chapter 1 Introduction Chapter 2 Methodology Chapter 3 Theoretical background Chapter 6 Case Creekboard-Giraffe Chapter 4 Case Riverboard-Mouse Chapter 5 Case Brookboard-Elephant Chapter 7 Case Lakeboard-Panda Chapter 8 Cross-case analysis: a static perspective Chapter 10 Theory building through simulation Chapter 11 Conclusion and reflections Within-case analysis Chapter 9 Cross-case analysis: a dynamic perspective

(24)

11

Research Methodology

This chapter presents Engaged Scholarship (ES) as the main research strategy. Section 2.1 introduces the motivation for using ES as the research strategy and explains what ES is. Section 2.2 describes the research process, which gives more detailed information on how the engaged scholarship strategy was applied to this research. Section 2.3 discusses the problem formulation, which explains the context of this research (i.e. Dutch water sector) and the challenges that public organizations (i.e. Dutch water authorities) face while engaging in PPPs. Section 2.4 presents a multi-method research design including a multiple-case study and a simulation. Section 2.5 discusses the research credibility consisting of the validity and reliability of this research.

2.1 Research strategy: Engaged Scholarship

Motivation

This research arose out of the project PURGATIO (Prestatiecontract- Uitvoerbaarheid voor ReiniGing Afvalwater Techniek In Outsourcingscontext), which involved multiple stakeholders, namely Tilburg University, World Class Maintenance, and four Dutch water authorities. The objective was to gain more understanding of how to successfully apply Performance-Based Contracts (PBCs) in outsourced renovation projects involving waste-water treatment plants in PPPs (see Section 2.3.1). For many reasons, this was a complex research setting, for example the involvement of multiple stakeholders, complex contracts, innovative processes, and long-term relationships. The intended research outcome from industrial side was to provide them with relevant practical knowledge for addressing real-world problems. The purpose was also to advance theoretical knowledge in the discipline in a rigorous way.

(25)

12

Ven (2007) as a participative form of research for obtaining the advice and perspectives of key stakeholders (e.g. researchers and practitioners) with a view to understanding complex problems or phenomena. As Shepherd & Suddaby (2017) point out, ES is likely to be most useful when associated projects involve complex real-world problems, a collaborative learning environment, an extended duration and multiple frames of reference. The problem-driven research, like this research, requires researchers to be engaged with the practitioners performing their activities, to be open to new experiences, and to be self-reflective on their engaged scholarship role (Van de Ven & Johnson, 2006). By doing this, researchers get a step closer to the challenge of bridging the gap between practice and theory (Hodgkinson, Herriot, & Anderson, 2001). This gap is further discussed in Section 2.1.2. Collaborating with practitioners throughout the process, researchers are able to formulate problems grounded in the experiences of those engaged a real-world problem, whose solution can contribute to both academic and practitioner knowledge (Van de Ven, 2007).

Gap between practice and theory

(26)

13

theories that often take the form of formal logical principles or rules involving causal relationships. In this line of reasoning, it is assumed that the relationship between knowledge of theory and practice entails a literal transfer or translation of one into the other.

Bridging the gap

Van de Ven (2007) takes a pluralist view of science and practice as representing distinct kinds of knowledge that are not opposites nor alternatives, but rather provide complementary insights for understanding reality. Van de Ven (2007) suggests that each form of knowledge is partial. The strengths of one form of knowledge tend to be the weaknesses of the other. Different perspectives and kinds of knowledge are incomplete and involve bias with respect to any complex problem. For that reason, a pluralist approach to knowledge co-production is needed. Van de Ven (2007) also argues that many social researches tend to reflect an unengaged process of inquiry, meaning that researchers study a research question by themselves, without interacting with other stakeholders who could actually help to shed insightful light on the problem domain under investigation.

To produce knowledge that can get over the dual hurdles of relevance and rigor, a deeper form of research that engages both academics and practitioners is needed (Hodgkinson et al., 2001). In other words, a mode of inquiry is needed that converts the information obtained by scholars in interaction with practitioners (and other stakeholders) into actions relevant for a given professional domain Van de Ven (2007).

Engaged scholarship

(27)

14

This strategy emphasizes that the advice and perspectives of key stakeholders (researchers, users, clients, sponsors, and practitioners) are required for the study of a complex social phenomenon, because the research problem can only be studied thoroughly by exploiting the pooled knowledge of scholars and other stakeholders. Engaged scholarship has the explicit intention of knowledge ‘co-production’ (Gibbons et al., 1994), as the knowledge produced in the study is more penetrating and insightful than that gained by either scholars or practitioners working alone on the problem (Van de Ven, 2007). Scholarship means not only engaging in original research but also looking for connections, building bridges between theory and practice, and communicating one’s knowledge effectively (Boyer, 1990: 16). Engagement implies that, while performing each step of the research process, scholars step outside of themselves to obtain and be informed by the interpretations of others: problem formulation, theory building, research design, and problem solving (Figure 2.1).

Theory Building

Create, elaborate and justify a theory by abduction, deduction or induction Engage knowledge experts in relevant disciplines

Problem Formulation

Situate, ground, diagnose and infer the problem

Engage those who experience and know the problem

Research Design

Develop model to study theory Engage experts and people providing access to information

Problem Solving

Communicate, interpret and negotiate finding with intended audience Engage intended audience to interpret meanings and uses

Th_eo ry B uild ing Rese arch Des ign Pro ble_m So lvin_g Prob lem Form ulat ion

Iterate and Fit

Solution Theory

Model

Reality

Figure 2.1 Engaged scholarship model, adapted from Van de Ven (2007)

(28)

15

experience and understand the problem, as well as reviewing the literature on the prevalence and boundary conditions of the problem.

In the theory building stage, researchers need to create, elaborate and justify a theory by abductive, deductive and inductive reasoning. Developing this theory and its plausible alternatives requires conversations with knowledge experts from the relevant disciplines and functions that have addressed the problem, as well as a review of relevant literature.

Moreover, researchers need to develop a variance or process model for empirically examining the alternative theories in the research design. Doing this well typically requires getting advice from technical experts in research methodology and the people who can provide access to data, and of course, the respondents or informants of information.

(29)

16

2.2 Research process

Problem formulation began in September 2013. Initial meetings were arranged in each of the public organizations (i.e. Dutch water authorities) between researchers from the university and managers from public organizations. I also joined many guided site tours to learn the business context of these public organizations. During the first year of research, several problem formulation sessions were conducted, followed up by group meetings, in order to identify a research problem that combines both parties’ interests. A literature review was carried out in parallel with problem formulation. This helped to identify theories relevant to the research problem and was useful in refining the research question. The literature review was mostly done between 2013 and 2015, and theory development between 2016 and 2018.

After this, data was collected and analyzed from December 2013 until December 2018. Empirical data were collected via interviews, field observation, and archive documents. The data was continuously analyzed while new data was collected, and this helped refine research questions and the direction of theory building, which in turn helped decision making regarding new data collection.

(30)

17

review that helped to develop and refine the research question is discussed in Chapter 3, the theory development is discussed in Chapter 10, and knowledge gained from this research is discussed in Chapter 11.

2.3 Problem formulation

This section introduces the Dutch water authorities (DWAs) and the challenges that they face while engaging in public private relationships. It concludes with the research problem which was discovered after the engagement with managers from DWAs.

Dutch water authorities

The water authorities are the oldest form of democratic government in the Netherlands. Their very first existence traces back to the 13th century and they have been vested in the Dutch constitution since 1848. They have three fundamental attributes:

 Their top management is chosen via public election every four years.

 They are empowered to levy their own taxes, meaning that they represent a separate public governance level, alongside the Dutch national government, province, and municipality administrations (Van den Oever, 2017). For example, the total expenditure of the water authorities was 2.9 billion euros in 2016, while a household that owned its own home had to pay an average of 805 euros in rates and taxes (Havekes et al., 2017).

(31)

18

up the top management team which meets approximately every two weeks (Van den Oever, 2017).

2.3.1.1 Responsibilities and the importance of Dutch water authorities The water authorities are responsible for water management on a regional and local level. The term ‘water management’ can be described as the part of public welfare that relates to flood protection, water management (surface water and groundwater in terms of both quantity and quality, including waste-water management) and the waterways. It also focuses on the habitability and usability of the soil and on the protection and improvement of the living environment (Havekes et al., 2017). There are 21 water authorities in the Netherlands by 2017. The infrastructure, such as flood defenses, pumping stations, flood barriers, canals and ditches (see Figure 2.2 for facts and figures) that they build and manage are vital in safeguarding a liveable, accessible and safe Netherlands. Given the special geographic situation in the Netherlands, without proper and continual maintenance of this infrastructure the safety of ten million Dutch citizens would be in immediate danger, as more than fifty percent of the Netherlands would be under water if the water management was not up to standard (Havekes et al., 2017).

Number of water authorities 21

Number of employees 11,250

Length of primary flood defenses being managed 3,600 kilometers

Length of other flood defenses 14,100 kilometers

Length of managed watercourses 230,000 kilometers

Number of pumping stations 3,550

Length of managed roads 7,500 kilometers

Number of wastewater treatment plants 335

Volume of wastewater treated 2 billion m3

(32)

19 2.3.1.2 Waste-water treatment

This research focuses on the Dutch water authorities’ responsibilities for waste-water treatment. A major part of their task is the construction, operation and maintenance of waste-water treatment plants where households and industrial wastewater is purified. Each year DWAs invest approximately 300 million euros in total in the field of waste- water management (Groot, Afrian, Suiskind, & Vrolijk, 2013). The waste-water treatment is a complex process involving two main processes, namely water purification and sludge treatment. The water purification process consists of screening, pre-sedimentation, aeration and final sedimentation. After the water purification process the water is free from harmful substances and chemicals and is therefore allowed to be released to rivers or the sea. The sludge (the residue of the water purification process) is handled in the sludge treatment process, which is composed of pre-dewatering, digestion and final dewatering. The byproduct of sludge treatment is biogas, which is used to produce electricity and heat. The final product of dewatered sludge (disposal sludge) is transported to the incineration plant for final treatment (see more details in Appendix 1).

2.3.1.3 PPPs in Dutch water sector

(33)

20

the design, build, operate and maintain (DBOM) of the installations that meet pre-set functional specifications.

Challenges

The abovementioned capital-intensive infrastructural installation characteristically has a long-life cycle, and this creates challenges for the DWAs in terms of coming up with appropriate PBCs for outsourcing, in particular for projects that include design, building, operation and maintenance. The contracts typically have a time horizon of 5-15 years (see more characteristics in Appendix 2), while the life cycles of installations encompass several decades. The challenge, then, is to build into the contract incentives that will lead contractors to perform well, not only in the short term but also in the long run. An intensively used installation is very unlikely to stand still without good maintenance. In that case, DWAs have to pay a huge bill due to extra cost.

(34)

21

In addition, due to the public function of DWAs, the outsourcing is also characterized by political aspects, such as European procurement rules and environmental rules. Given these challenges as a result of complexity in technology, contract and environment, DWAs are required to find a new way of managing contract and public – private relationships. However, experience to date with these performance-based contracts is still rather limited, causing unsatisfied outsourcing performance, either financially, technologically or relationally.

Research problems

In the problem formulation stage, I conducted nine interviews with managers from DWAs, aiming to identify a research problem that is interesting from both a practical and academic perspective. After the engagement, three main problems that middle management have encountered or witnessed emerged.

 The first one related to contract management, which was “how to properly manage contracts in general?” In particular, how to adapt to changing environments and cope with unexpected problems, given that the contracts with contractors have been signed.

 The second problem concerned relationship management, which was mainly about mutual understanding and trust between partners. The following questions were raised by managers: How to build mutual trust with a contractor during the whole contracting process? Why do people not understand each other and how can this be improved?

(35)

22

outcome of outsourcing projects, they were eager to discover what has caused the unfavorable results.

After two group feedback sessions with managers, all stakeholders (the author and managers from Dutch water authorities) realized that there was a link between the third problem and the first & second problems, since contract management and relationship management might serve individually or jointly as underlying mechanisms for the success or the failure of an outsourcing project in a PPP. We therefore decided to investigate the third problem, while taking the perspectives of contract and relational management. Thus, the research focuses on finding out the root causes of the success or the failure of an outsourced renovation project in Dutch water authorities.

2.4 Research design

This section first presents two basic approaches that are commonly applied in empirical studies, variance and process research design. Then it explains why the process research design was chosen and how it was designed, including data collection and data analysis.

Variance and process research design

(36)

23

or event-driven explanation of the temporal order and sequence in which a set of events occur based on a story or narrative (Bruner, 1991).

With regard to causality, “what” questions entail evidence of co-variation, temporal precedence, and absence of spurious associations between the independent and dependent variables (Blalock, 1972), while “how” questions require narratives that indicate an observed sequence of events with an underlying generative mechanism explaining the causes of real life events or contingencies (Bruner, 1991; Tsoukas, 1989). As a conclusion (Figure 2.3), in general terms, a variance theory explains change in terms of relationships among independent variables and dependent variables, whilst a process theory explains how a sequence of events leads to some outcome (Mohr, 1982). To elaborate more on process theory, it explains development in terms of the order in which things occur and the stage in the process at which they occur. In narrative methods the plot in the story itself is the generative mechanism. Process models strive to build theory that depends on versatility, defined as “the degree to which it can encompass a broad domain of developmental patterns without modification of its essential character” (Poole et al., 2000: 43). A versatile process explanation can stretch or shrink to fit specific cases that may differ in their tempo and time span (Van de Ven, 2007).

(37)

24

Process research design

The process model is adopted in this research, because the goal is to investigate a “how” question (i.e. how do two governance mechanism and performance affect each other over time?). The process perspective focuses on actors, temporal orders and the sequence of critical events that explain how things change over time (Langley, 1999; Langley, Smallman, Tsoukas, & Van de Ven, 2013; Pettigrew, 1997; Van de Ven, 1992). Studying governance and performance as processes through the analysis of actors (i.e. public organizations and private companies) and critical events, could allow us to gain an enhanced understanding of how these two – governance and performance – interact with each other.

Following Chandrasekaran et al. (2016) and Gray, Esenduran, Rungtusanatham, & Skowronski (2017), I applied a multi-method research approach (i.e. a combination of a multiple case study and a system dynamics simulation method) to explore such a complex phenomenon as Public – Private Partnerships (see Figure 2.4 ).

(38)

25 Field Observations: Projects with

Performance based contracts/DBMO contracts in a Public-Private Partnership setting experienced performance shortfalls.

Case Analyses: Within and cross-case

analyses showed contractual and relational strategies in managing performance shortfalls resulted in different project outcomes.

Theory Development: A cross-case comparison identifies a general mechanism of how contractual and relational strategies affect performance dynamics.

Theory Refinement: Develop a contingency

approach on the effectiveness of contractual strategies and relational strategies.

Simulation Analyses: A system dynamics model is able to replicate performance outcomes under different strategies in real cases and helps identify boundary conditions of strategies in dealing with performance shortfalls.

Theory Augmentation: Iterating between case analysis and the literature gives rise to the relational strategies, rather than contractual strategies, in dealing with performance shortfall in complex projects.

Method 1: Multiple Case Studies Method 2: System Dynamics

Simulation

Figure 2.4 A multi-method study design adapted from Chandrasekaran et al. (2016)

Multiple case study

(39)

26

other was in a relatively bad situation. The same applies to the other two cases. The case from Riverboard (see Chapter 4) was a relatively successful case at the time we started this research, as the two parties were able to find the technical solution to the unexpected technical problems. While in the case of Creekboard (Chapter 6) the two parties were already engaged in contractual discussions, in the case of Brookboard, the two parties were fighting in court when case was selected, and in the case of Lakeboard, the two parties were in a good relationship at that point in time.

The unit of analysis refers to the entities (individuals, collectives, or objects) being studied, which can be identified as who or what is to be described or analyzed in a study (Van de Ven, 2007). In other words, it defines what the case is and how it is related to the way the research questions have been defined (Yin, 2014). The main unit of analysis in this research is a dyadic relationship between public organizations (Dutch water authorities) and their contractors (engineering and construction companies) in Public – Private Partnerships (PPPs) in the Netherlands. In view of the objective of this research, this unit of analysis enables an investigation into the dynamic interaction process between two types of governance and performance in a PPP.

Data collection

(40)

27 2.4.4.1 Time horizon

The research started in September 2013 and lasted for more than four years. As the case projects had started prior to data collection, some of the data is retrospective in nature, while the majority is collected in real time. Data collection took place between December 2013 and June 2018. Data collection in the case of Riverboard – Mouse and the case of Creekboard – Giraffe was stopped in June 2016, as both cases were already at the operation and maintenance stage and in both cases the two parties had been able to overcome the performance shortfall and improve their relationship. The development of relationship tended to fall in the positive direction and stay stable. Data colletion was stopped in the case of Brookboard – Elephant in January 2017, as both parties terminated the contract in December 2016. Although the case of Lakeboard – Panda is ongoing and still at the construction stage, data collection was stopped in June 2018, mainly because the end of the research period was approaching.

2.4.4.2 Semi-structured interviews

(41)

inter-28

organizational project team meetings. The data collection process eventually resulted in total of 74 interviews with average duration of approximate one hour per interview. The interviews were later transcribed and yielded 633 pages of transcription. Appendix 4 shows more information about interviewees and when the interviews have been conducted. Table 2.1 presents an overview of the collected data. Cross-validation of interviews (Gioia & Thomas, 1996) took place by summarizing the results from the previous interview at the beginning of each new interview. The transcriptions were also sent back to interviewees for verification.

Table 2.1 Overview of data

Overview of data

Contract Interview Minutes

Contract

supplement Performance Others

Cases NO. Pages NO.

Duration

(h) Pages NO. Pages NO. Pages NO.

Pag es NO . Pages Brookboard – Elephant 1 72 20 18 181 32 338 5 302 12 20 17 202 Riverboard – Mouse 3 78 22 30 204 58 241 10 188 13 50 24 123 Creekboard – Giraffe 2 12 22 17 159 48 365 18 999 33 323 14 92 Lakeboard – Panda 1 38 10 10 89 1 6 7 248 0 0 11 208 Total 7 200 74 75 633 139 950 40 1489 58 393 66 625

Total number of Documents 384

Total number of Pages 4290

2.4.4.3 Observation and other documents

(42)

29

Akkermans, Sengupta, & Wassenhove, 2013), and the study of other relevant data sources including the seven contracts (200 pages), 40 contract supplements (1489 pages), 139 minutes of meetings (950 pages), 58 performance documents (393 pages) and 66 other documents (e.g. website announcements, newspaper articles; 625 pages). In total 384 documents (4290 pages) were collected. This approach allows perceptual data from interviews to complement objective data from other sources, and helps to reduce retrospective bias (Langley, 1999) as well as informant bias (Ketokivi & Schroeder, 2004).

Data analysis

2.4.5.1 Data analysis process

The qualitative data was analyzed at both within case and across case levels. The process of data analysis is visualized in Figure 2.6. As Miles et al. (2014) suggest, data analysis consists of three iterative steps: data condensation, data display and drawing & verifying conclusions. Data collection and data analysis can be considered as interconnected and overlapping processes rather than independent and separate ones. The three steps of data analysis will be further explained in the sub-sections on within case analysis and cross case analysis.

Data collection Data condensation Data display Drawing & verifying conclusions

(43)

30 2.4.5.2 Within case analysis

Data condensation refers to the process of selecting, focusing, simplifying, abstracting and transforming the data that appear in the full body of written up field notes, interview transcripts, documents and other empirical materials (Miles et al., 2014). Coding is primarily used to condense data by categorizing similar data chunks so that a researcher can quickly find, pull out and cluster the segments that relate to a particular construct or theme (Miles et al., 2014). Following the general guidelines of coding suggested by Miles et al. (2014), first and then second cycle coding (pattern coding) was applied. First cycle coding is a way of initially summarizing segments of data, while pattern coding, as a second cycle method, is a way of grouping those summaries into a smaller number of categories, themes or constructs.

More specifically, the principal investigator first coded the data using NVIVO (see example of coding in figure 2.7), condensing data through an open coding process by using the interviewees’ vocabulary, such as “contract,” “relationship,” “project,” and “contracting process.” These codes also had several “sub codes” (e.g. “trust” being a sub code of “relationship”) and associated “dimensions” (e.g. “competence trust” and “goodwill trust”). The initial coding list is shown in Appendix 5.

(44)

31

(45)

32

performance shortfall happened, how the two parties in the PPP responded to it and how this interaction affected the project performance and relational outcome. At the end of each narrative, the narrative is reflected upon and explicit connections made to the key concepts, namely project complexity, performance shortfall, strategies, project performance, and relational outcome (see their definitions in Appendix 6), under a static perspective. Subsequently, a dynamic perspective is provided by displaying an underlying causal mechanism in a causal chain (Figure 4.2, 5.2, 6.2 and 7.2). This is a researcher-constructed linear display of events, action and/ states that suggests a plausible sequence of causes and effects. Such a chain helps a researcher to lay out explicitly what may be causing certain phenomena Miles et al. (2014).

2.4.5.3 Cross-case analysis

Just as in the within-case analysis, the four cases are analyzed from a static and a dynamic perspective in the cross-case analysis. In the static perspective (see Chapter 8), the commonalities and differences between four cases are compared on five key concepts, namely project complexity, performance shortfall, strategies, project performance, and relational outcome. The aim is to investigate when and why a performance shortfall occurs and what the relationships are between contractual and relational strategies in dealing with a performance shortfall and project outcomes. In the dynamic perspective (Chapter 9), the four cases are compared on the dynamic processes between contractual/relational strategies and project outcomes. The objective is to identify a general mechanism of how strategies affect performance dynamics. Specifically, a causal loop diagram that depicts hypothesized causal structure of a PPP under investigation will be developed.

(46)

33

interrelatedness between different variables in a system, in this case a public – private relationship. It shows the causal links among variables with arrows from a cause to an effect. The important feedback loops are also identified in the diagram. A positive link indicates a positive relation and a negative one indicates a negative relation. A positive causal link means a decrease/increase in a variable resulting in a decrease/increase in the other variable. In other words, two variables change in the same direction. While a negative causal link means the two variables change in opposite directions, i.e. if one variable starts increases, the other variable decreases and vice versa. A feedback loop is one of the important features in a causal loop diagram, which is either defined as a positive or negative loop. In positive causal loops, the feedback is self-reinforcing and amplifying, while in a negative causal loop, the feedback is dampening (Sterman, 2000). As a conclusion, by comparing four cases, a general causal loop diagram is developed, and has the ability to explain why and how two governance mechanism led to different performance dynamics. Examples of how causal relationships between variables is collected, and how a causal network is constructed using a causal loop diagram can be found in Appendix 7.

Simulation

In this research, a simulation model was built based on the general causal loop diagram discussed above. As Davis et al. (2007) point out, simulation has become an increasingly significant methodology in theory building in management literature. The purpose of using a simulation model is to refine the theory that developed after the cross-case analysis. The rest of this section explains why a simulation approach is used and why the system dynamics simulation modelling is chosen.

(47)

34

It is particularly useful to apply simulation after the theory-creating researches using methods such as inductive multiple case studies (Eisenhardt, 1989). This is also the case in this research, since simulation enables the elaboration of rough, basic or simple theory that is derived from inductive cases into logically precise and comprehensive theory. Simple theory is an undeveloped theory that has only a few constructs and related propositions with modest empirical or analytic grounding such that the propositions are in all likelihood correct but are currently limited by weak conceptualization of constructs, few propositions linking these constructs together, and/or rough underlying theoretical logic. Thus the theory developed through simulation can then be effectively examined further by using deductive logic and empirical evidence. To be more specific, through computational experimentation simulation allows the enhancement of a theoretical precision and related internal validity, and enables theoretical elaboration and exploration. In particular, simulation relies on some theoretical understanding of the focal phenomena in order to construct a computational representation. And simulation also depends on an incomplete theoretical understanding such that fresh theoretical insights are possible from the precision that simulation enforces and the experimentation that simulation enables. Furthermore, simulation is particularly useful when the theoretical focus is longitudinal, nonlinear, or processual (Davis et al., 2007), as is the case in this research, since in these situations, simulation is likely to reveal non-intuitive elaborations of simple theory that are difficult to uncover using other methods.

System Dynamics as simulation approach

(48)

35

(49)

36

2.5 Research credibility: validity and reliability

Yin (2014) suggested four criteria to assess the credibility of a qualitative case research design: construct validity, internal validity, external validity, and reliability.

Construct validity

Construct validity refers to the extent to which correct operational measures are established for constructs under study (Yin, 2014) and threatened by potential research subjectivity. To meet the requirement of construct validity, as recommended by Yin (2014), the main topics (e.g. contractual governance and relational governance) were defined in terms of specific concepts and operational measures (e.g. contractual control, self-serving response, accommodative response and information sharing) were identified to match the concepts by citing published studies that have the same matches (see Appendix 5).

To mitigate the negative effect of researchers’ subjectivity, the three methods proposed by Ellram (1996) are followed, namely multiple data sources, establishing & maintaining a chain of evidence, and having key informants review the reports.

 Three data sources of evidence were applied: periodic interviews, field observations and archive documents (e.g. contracts, minutes of meetings, and media publications), to confirm the validity of information being collected.  A computer-assisted data base was created using qualitative data analysis

(50)

37

any evidence from initial research questions to ultimate case study conclusions, or the other way around, can be traced.

 Transcripts of the interviews were reviewed by interviewees after each interview to avoid wrong interpretation by the interviewer. Reports were presented in group meetings periodically and feedback from managers incorporated into the revised reports.

Figure 2.7 A screenshot of Nvivo 10

Internal validity

(51)

38

theoretical statement, the statement revised, and the revision compared to the findings of the second case and so on.

External validity

External validity is the extent to which a study’s findings can be generalized (Yin, 2014). The lack of generalizability has been a major criticism of qualitative case study, which can be improved by using replication strategy in multiple case studies. In contrast to statistical generalization in quantitative research, qualitative case study focuses on analytical generalization. More specifically, a statistical inference is based on a random sample from a certain population, and the statement obtained is in principle valid for the whole population. While analytical generalization focuses on generalizing a particular set of results to some broader theory (Yin, 2014). The aim of this research is to extend the findings from the specific setting of PPPs between Dutch water authorities and their contractors to PPPs in different sectors or industries.

Reliability

(52)

39

Theoretical background

This chapter shows the theoretical background for the research. Section 3.1 reviews the literature of two exchange governances, namely contractual and relational governances. After explaining the definitions, types and dimensions, and functions of the two governances, the chapter discusses the performance outcomes that these two governances might bring. The chapter then presents the debate on the interplay between these two governances, namely their mutual relationship and their joint impact on performance. Section 3.1 concludes by pointing out the theoretical gap in the existing literature about the two governances. Under the theoretical framework of the two exchange governances, section 3.2 examines the literature on strategies for dealing with performance shortfall. The aim is to further refine the research question in the specific context of this research on public – private partnerships. Section 3.3 discusses the results of the literature review and concludes with the refined research question.

3.1 Two exchange governances

Two distinct types of governance mechanisms for supporting and managing Inter-Organizational Relationships (IORs) have been studied in the literature (Griffith & Myers, 2005; Rousseau, Sitkin, Burt, & Camerer, 1998; Williamson & Ouchi, 1980; Zaheer & Venkatraman, 1995). One of these is contractual governance, emphasizing the importance of formal rules and contracts between organizations for governing the exchange. The other is relational governance, relying on relational norms and trust to coordinate the exchange relationship, mitigate hazards associated with uncertainty and transaction-specific investment, and allow adaptation to unexpected future contingencies (Cao and Lumineau, 2015).

(53)

40

governances in IORs has been a hot topic in management literature. Nevertheless, to date there is little consensus on the relationship between these two governances and their impacts on performance. On the one hand, many scholars (Huber et al., 2013; Li et al., 2010; Lui and Ngo, 2004) argue that contractual and relational governance substitute for each other. In other words, the use of one type of governance decreases the use of or the benefits of using the other one (Cao and Lumineau, 2015). On the other hand, other scholars (Lui et al., 2009; Poppo and Zenger, 2002) contend that the two types of governance are complementary, meaning that the use of one increases the use of or the benefits of using the other one.

Contractual governance

3.1.1.1 Definition and dimension

(54)

41

formal and written contract which explicitly stipulates the responsibilities and obligations of each party (Ryall & Sampson, 2009).

Contracts, as the main type of contractual governance, are written agreements that are legally binding (Klein-Woolthuis, Hillebrand, & Nooteboom, 2005). Recently, Performance Based Contracts (PBCs) have been increasingly applied in practice, in both public (e.g. government, healthcare) and private (e.g. manufacturing and service industries) sectors (Hooper, 2008; Hypko et al., 2010). PBCs have also attracted much academic attention (Selviaridis & Wynstra, 2015). In short, PBC is considered a contractual approach that links payment to contractor’s performance. Moreover, it places emphasis on specification and evaluation of outputs or outcomes rather than required inputs, activities or processes (Martin, 2007). In other words, PBC underlines the outcomes of exchange rather than how to organise transactions or which resources to use (i.e., it allows the supplier some freedom). As such, a PBC clearly separates the client’s expectations (i.e., the performance goal) from the partner’s implementation (i.e., how it is achieved) (Kim, Cohen, & Netessine, 2007). PBCs are increasingly applied in the specific setting of this research (i.e. Dutch water authorities) and are a type of contractual governance predominantly used in the context of partnering with an organization that delivers services (Sumo, van der Valk, van Weele, & Duysters, 2016). This construct is further elaborated in the rest of this section.

(55)

42

reported and monitored. After the evaluation of performance, payment should be effected and corrective activities should be planned (Mohammadnezhad-Shourkaei, Abiri-Jahromi, & Fotuhi-Firuzabad, 2011; Schulz, Wu, & Chow, 2010). Moreover, the incentives dimension is considered as the structure of financial & non-financial incentives and their impact on supplier behavior. It stresses that the incentives structure that is designed and its intensity influence supplier behavior in terms of incentive alignment (Huntington, Zaky, Shawky, Fattah, & El-Hadary, 2010). This refers to the process of sharing costs, risks, and benefits between partners (Simatupang & Sridharan, 2002). If the incentive is aligned, supplier effort is directed towards improving performance, otherwise it might create unintended consequences and induce supplier opportunism (Mcdonald & Roland, 2009). The risk dimension refers to the allocation of financial and operational risk. The main motivation of buyers to apply PBC is to transfer risks to suppliers by linking reward to performance (Nalli, Scanlon, & Libby, 2007). However, suppliers can be risk averse, if they sense limited control of performance that depends on outcome clarity, measurability and intensity, and perceived capacity to manage related risks (Haghifam & Abedi, 2013).

3.1.1.2 Functions

(56)

43

Safeguard mechanism

Through specifying each party’s rights and duties in the contract, contractual governance may mitigate ex ante and ex post risk of opportunism and thereby safeguard the exchange between partners (Schepker et al., 2014; Williamson, 1985). Prior literature has regarded contractual governance as an important way of controlling exchange hazards, by minimizing cost and performance losses resulting from hazards (Joskow, 1988; Laura Poppo & Zenger, 2002; Weber & Mayer, 2011).

(57)

44

instance, Parkhe (1993) points out that formal contracts may stipulate penalties that change the pay-off structure by increasing the cost for opportunistic behaviors.

Coordination mechanism

(58)

45

Adaptation mechanism

It is also argued by Schepker et al. (2014) that contractual governance can serve an adaptation function in an uncertain environment. Providing evidence that adaptation functions of contractual governance, Argyres et al. (2007) discover that as firms contract over time, they include more contingency planning in the contracts. At the same time, Mayer and Argyres (2004) also find that contractual terms are added to future contracts as contingencies arise. The mechanism for adaptation is to define mutually agreed tolerance zones between parties for unexpected contingencies or to develop procedures to follow when changes start to appear. In other words, to function as an adaptation mechanism, a contract should contain provisions that define remedies for foreseeable contingencies, or specify processes for resolving unforeseeable outcomes (Poppo and Zenger, 2002). For instance, Crocker and Reynolds (1993) find that contracts with mutual agreed payoff schemes that depend on dynamic environmental circumstances help parties to cope with price fluctuation.

Relational governance

3.1.2.1 Definition and dimensions

(59)

46

considered as an unwritten mechanism for governing exchange, such as an interest in preserving the value of ongoing relationships that prevents opportunistic behavior. Though relational governance has been associated with different titles and definitions, it has been commonly associated with being informal and social in comparison with formal contractual governance (Vandaele et al., 2007). Relational governance is based on values and shared processes in the exchange relationship and incorporates trust and commitment (Lui & Ngo, 2004), relational capital (Kale, Dyer, & Singh, 2002), information sharing routines (Poppo et al., 2008), and informal exchange (Cook & Emerson, 1978). In line with Cao and Lumineau (2015), this study defines relational governance as the extent to which an IOR is governed by informal and social relations and shared norms (Laura Poppo et al., 2008; Zhou & Xu, 2012).

(60)

47

Relational norms relates to shared expectations about the behavior of each party in IORs (Cannon, Achrol, & Gundlach, 2000; Heide & John, 1992). The main dimension of relational norms are flexibility, information exchange, and solidarity (Griffith & Myers, 2005). More specifically, flexibility is conceptualized as willingness to adapt to one’s partner within the context of an exchange relationship, while information exchange defines a bilateral expectation that partners will proactively provide information useful to their partner in support of the ongoing relationship. Solidarity refers to the bilateral expectation that a high value is placed on the relationship (Heide, 1994; Heide & Miner, 1992; Laura Poppo & Zenger, 2002).

3.1.2.2 Functions

As discussed in Section 3.1.1, contractual governance may serve three types of function. Nevertheless, in the existing literature no study has yet specifically associated relational governance to function. This study argues that relational governance can also be applied as a safeguard for exchange, to enhance coordination between partners, and to promote adaptation when unexpected events occur.

Safeguard mechanism

Relational governance highlights the role of trust and social norms in IORs (Griffith & Myers, 2005; Gulati, 1995). On the one hand, Ring and Ven, van de (1992, 1994) suggest that as a key element in relational governance, trust helps to reduce the risk of opportunistic behavior. When a high level of mutual trust exists in a relationship, both parties share mutual confidence or belief that both parties will not exploit any adverse situation (Barney & Hansen, 1994) and they are more likely to consider their partner’s interests rather than just their own (Lui et al., 2009). Thus, trust and its underlying normative implication operates as a self-enforcing safeguard in preventing opportunistic behaviors (Laura Poppo & Zenger, 2002).

When performance shortfall arises, contract or trust? A multi-method study of the impact of contractual and relational governances on performance in public – private partnerships

When performance shortfall arises, contract or trust?

Acknowledgement

Table of Contents

Introduction

1.1 Research context

The infrastructure sector

Public – private partnerships (PPPs)

1.2 Research problem: performance shortfalls

1.3 Research question

1.4 Structure of the dissertation

Research Methodology

2.1 Research strategy: Engaged Scholarship

Motivation

Gap between practice and theory

Bridging the gap

Engaged scholarship

2.2 Research process

2.3 Problem formulation

Dutch water authorities

Challenges

Research problems

2.4 Research design

Variance and process research design

Process research design

Multiple case study

Data collection

Data analysis

Simulation

2.5 Research credibility: validity and reliability

Construct validity

Internal validity

External validity

Reliability

Theoretical background

3.1 Two exchange governances

Contractual governance

Relational governance